The objective of this grant application is to use genomic technology to identify a novel cluster of biomarkers expressed by peripheral blood leukocytes (PBL) that will be clinically useful in the identification of patients with tibio-femoral knee osteoarthritis (OA), and to identify those patients at risk for disease progression. Preliminary data, funded for the past year under the OA Biomarkers Network, indicates that PBL derived from a subset OA patients exhibit a "cytokine activation profile", suggesting leukocyte exposure to stimuli during the course of perfusion of diseases joint tissues. To pursue these findings, the grant will have three central aims: In Specific Aim 1 we will determine whether PBL gene expression patterns distinguish early OA (OA-E, K-L grade 2), advanced OA (OA-A, K-L grade 3) and healthy controls as defined by semi-flexed, fluoroscopically-positioned knee radiographs. For these 3 cohorts, overall and pairwise comparisons of PBL gene expression profiles will be analyzed in order to identify a distinctive profile which discriminates the following: 1) OA versus normal, and 2) early versus advanced stage OA. We will validate PBL gene expression profiles of knee OA in an independent population, the Duke knee OA cohort, in collaboration with Dr. Virginia Kraus. In Specific Aim 2 we will examine the association of PBL gene expression profiles with findings on MRI and bone scintigraphy. We will perform MRI studies using a clinical 3.0 T MRI system on 180 patients at NYU-HJD, who will be entered into a two year longitudinal cohort study. We will perform semiquantitative analyses of articular cartilage integrity, subchondral marrow lesions, synovitis/effusion, marginal osteophytes, assessing both global and individual compartment scores. In collaboration with Dr. Kraus we also examine the association of PBL gene expression profiles with signal knee "synovitis" and with "total body OA burden", as defined by early and late phase semi-quantitative bone scintigraphy, respectively. We will develop a multivariate model of OA radiographic stage to determine the independent associations of PBL gene expression profiles with other risk factors including BMI, COMP, bone scintigraphy and varus/valgus deformities. Specific Aim 3 will be a 2 year longitudinal to determine whether specific PBL gene expression patterns predict OA disease progression. We will assess OA patients with progression and without progression as determined by both MRI and semi-quantitative knee radiography. Imaging evaluations will occur at 0 and 24 months. In addition to MRI and x-ray, each subject will undergo: i) clinical evaluation including BMI determination, ii) pain and function assessment (complete WOMAC), iii) serum COMP. In addition, serum, plasma and urine will be collected for future biomarker determinations. We anticipate that these studies will enhance our understanding of the pathogenesis of OA and provide a combinatorial biomarker that may predict OA disease progression.